Remote-sensing Sensors Research Articles

Wrapper-Based Feature Subset Selection for Rapid Image Information Mining

In a disaster, there is a need for rapid image-information retrieval in real or near real time from vast amounts of data coming from multiple remote-sensing sensors. In general, image information mining (IIM) approaches produce enormous amounts of features that are computationally expensive and inefficient to process before the actual information discovery takes place. Also, it is complicated because the combination of the features has little relevance to the hypothesis space. Hence, selecting a relevant subset of features is necessary to overcome these problems and to provide an efficient representation of the target class. In this letter, we propose feature selection and feature transformations based on a wrapper-based genetic algorithm approach. A support vector machine classification is applied for generating predictive models for those land-cover classes that are important in a coastal disaster event. The proposed system, rapid IIM, is a region-based approach where, in lieu of the prevalent pixel-based methods, it localizes interesting zones and enables rapid querying. Results from this study indicate that selecting relevant feature subsets increases the rate of correctly identifying a semantic class and also enables this process with less number of features.

Remote Sensing Sensors and Applications in Environmental Resources Mapping and Modelling.

The history of remote sensing and development of different sensors for environmental and natural resources mapping and data acquisition is reviewed and reported. Application examples in urban studies, hydrological modeling such as land-cover and floodplain mapping, fractional vegetation cover and impervious surface area mapping, surface energy flux and micro-topography correlation studies is discussed. The review also discusses the use of remotely sensed-based rainfall and potential evapotranspiration for estimating crop water requirement satisfaction index and hence provides early warning information for growers. The review is not an exhaustive application of the remote sensing techniques rather a summary of some important applications in environmental studies and modeling.

Extraction rice-planted areas by RADARSAT data using neural networks

A classification technique using the neural networks has recently been developed. We apply a neural network of learning vector quantization (LVQ) to classify remote-sensing data, including microwave and optical sensors, for the estimation of a rice-planted area. The method has the capability of nonlinear discrimination, and the classification function is determined by learning. The satellite data were observed before and after planting rice in 1999. Three sets of RADARSAT and one set of SPOT/HRV data were used in Higashi–Hiroshima, Japan. Three RADARSAT images from April to June were used for this study. The LVQ classification was applied the RADARSAT and SPOT to evaluate the estimate of the area of planted-rice. The results show that the true production rate of the rice-planted area estimation of RADASAT by LVQ was approximately 60% compared with that of SPOT by LVQ. It is shown that the present method is much better than the SAR image classification by the maximum likelihood method.

Estimation of aboveground biomass in logged and primary lowland rainforests using 3-D photogrammetric analysis

We estimated the total aboveground tree biomass (TAGB) in an old-growth primary forest and in a regenerating forest that had been selectively logged in 1958, both within the tropical rainforest of the Pasoh Forest Reserve in Peninsular Malaysia. This was achieved by comparing aerial photographs with data obtained previously from destructive sampling in the same area. Aerial photographs were taken above the primary and logged forest plots in 1997. The heights of the canopy-forming trees were estimated in both plots by means of aerial triangulation and were regressed against the diameter at breast height (DBH) of the corresponding trees measured during ground surveys. The resulting allometric relationship between tree height and DBH let us estimate TAGB: in the primary forest, TAGB was 310 Mg ha −1, which was ca. 10–12% smaller than the value estimated by means of destructive sampling conducted in the 1970s. The estimated TAGB of the logged forest was 274 Mg ha −1, which was significantly smaller than that of the primary forest ( P < 0.05). We also measured canopy surface height in a 2.5 m grid system. We found that the mean canopy surface height (MCH) in every 20 m × 20 m subplot (0.04 ha) was significantly ( P < 0.0001) correlated with TAGB for that subplot. This suggests that the spatial variation of TAGB can be estimated using MCH values obtained from such a grid system, and that biomass levels can potentially be estimated by means of satellite remote sensing on larger scales, even for very dense tropical forests. We also found that digital reflectance values from Landsat Thematic Mapper (TM) images differed significantly between the logged and primary forests, and hypothesize that these differences relate to structural differences in the canopy surface. However, TAGB in both plots was poorly correlated with the Landsat reflectance values, suggesting the necessity of using an active remote-sensing sensor or a laser profiling system that can quantify changes in the forest's vertical structure or volume to estimate biomass and its variation in dense evergreen forests.

A system for monitoring NO 2 emissions from biomass burning by using GOME and ATSR-2 data

In this paper we propose a system for monitoring abnormal NO 2 emissions in the troposphere by using remote-sensing sensors. In particular, the system aims at estimating the amount of NO 2 resulting from biomass burning by exploiting the synergies between the GOME and the ATSR-2 sensors mounted on board of the ERS-2 satellite. Two different approaches to the estimation of NO 2 are proposed. The former, which is the simpler one, assumes a linear relationship between the GOME and ATSR-2 measurements and the NO 2 concentration. The latter exploits a nonlinear and nonparametric method based on a radial basis function (RBF) neural network. The architecture of such a network is defined in order to retrieve the values of NO 2 concentration on the basis of the GOME and ATSR-2 measurements, as well as of other ancillary input parameters. Experimental results, obtained on a real data set, confirm the effectiveness of the proposed system, which represents a promising tool for operational applications.

The effect of scintillation on the active microwave remote-sensing sensors

Abstract Recently, space-borne radar systems have been used to collect data for many different remote sensing applications. Some of the applications demand very high accuracy in the measured backscattering cross-sections. Microwave signals passing through the ionosphere may be, at times, affected by the scintillation effect, thus the backscattering cross-section. It is the purpose of this paper to discuss the gigahertz scintillation phenomenon and to assess the potential impact on active microwave observations.

Simultaneous Passive and Active Microwave Observations of Near-Shore Beaufort Sea Ice

Combined use of active and passive microwave imagery is the optimum way to observe the morphology and dynamics of near-shore ice. Such data are compared with the electrically scanning microwave radiometer imagery of the Nimbus-5 satellite. Data from the synthetic aperture radar, scanning multichannel microwave radiometer, Seasat-1, and Nimbus-7 can provide an increased understanding of Beaufort Sea nearshore ice. Background The current need for sea ice information has occurred at a time of rapid evolution of remote-sensing platforms and sensors. These timely technical advances are beginning to eliminate the observational barriers that have limited our knowledge of a natural phenomenon existing in areas that are dark and/or cloud-covered much of the year. Furthermore, as sea ice undergoes large spatial variations on short time scales, the new sensing techniques are acquiring for the first time the sequential synoptic observations needed for cause-and-effect studies. Considerable emphasis in recent years has been placed on the microwave remote sensing of sea ice because it offers the possibility of an all-weather day-or-night capability. Both passive and active microwave remote sensing techniques have been explored. An overview of these activities has been given in Ref. 1. Early work in this area used passive microwave techniques because such techniques were the first to be incorporated aboard earth-viewing satellites. The most important series of aircraft flights that demonstrated the feasibility and usefulness of ice observations by means of passive microwave sensors were those that occurred during the NASA Arctic Ice Dynamics Joint Experiment (AIDJEX). A series of three AIDJEX pilot field experiments were performed during the spring of 1970, 1971, and 1972 in the southern Beaufort Sea. During each of these experiments, the NASA CV-900 Galileo I performed a variety of flights ranging in altitude from 150 m to 11 km. A wide variety of visual and infrared sensors were operated in addition to an imaging radiometer operated at a wavelength of 1.55 cm and fixed-beam radiometers operated at wavelengths of 0.81, 2.8, 6.0, and 21 cm. The 1970 data showed that it was possible to distinguish sea ice from liquid water both through the clouds and in the dark. This finding was useful because it pointed the way to an "all-time" ability to observe leads and polynyas. These data also showed that strong microwave emissivity differences occur on the ice surface itself. However, the lack of sufficient ground-truth data prevented a determination of the reason for these differences. The ground-truth measurements and mesoscale microwave mosaic maps (10 000 km) acquired during the 1971 AIDJEX experiments allowed Gloersen et al. 3 to show that the observed emissivity differences of sea ice at a wavelength of 1.55 cm are associated with the age of the ice, with multiyear ice having low emissivities (cold brightness temperatures, 210 K) and first-year ice having higher emissivities (235 K). This was important because it suggested that passive microwave imagery could provide an all-time capability of distinguishing between old (thick) and new (thin) ice and of tracking ice motion as well as lead and polynya dynamics.